Growth Cycles, Network Effects, and Intersectoral Dependence: An Agent-Based Model and Simulation Analysis

An agent-based model of economic growth and technological change with network effects is proposed. The dynamics generated by network externalities are self-reinforcing and may bring about rapid growth, but will also for some time prevent further innovation. This circular pattern may appear in different economic sectors (or regions), may synchronize and resonate between sectors. This gives rise to growth waves on the macro-level and may be a novel approach to explain growth cycles. The paper uses an agent-based model for the study of single sector industry dynamics. This design is extended into a multi-sector version with an intersectoral effect on the network externality terms. The model is then simulated both with interconnected (with different network structures) and - as a control treatment - with isolated sectors. The emerging wave pattern on the macro-level is analyzed using both the autocorrelation spectrum and the frequency spectrum obtained with a fast Fourier transformation (FFT) of the simulation's output data.